One of the current difficulties in battling the destructive and costly wildfires is in obtaining up to date, accurate information of the fire's current location and intensity. Current methods relying primarily on satellite technology are too slow and inaccurate, therefore a better method is needed to help lower the destruction caused by wildfires and reduce the resources needed to battle t hem. This research proposes distributing a large number of cheap sensors across an area encompassed by wildfire capable of organizing themselves into an ad-hoc wireless sensor network to monitor the fire's current intensity and location. The majority of work performed pertaining to this research is in developing and analyzing the simulation tools needed to accurately test the wireless sensor network wildfire detection system. Commercially available software was used to generate realistic fires to test the system with, while custom software was developed to test how accurately randomly distributed sensors can predict a fire's outer perimeter intensity and location. The simulation results show this proposed method to be a promising solution to the current lack of information available for fighting wildfires. The predicted fireline is generated using a combination of algorithms to extract the most important information from the sensor nodes and generate a best guess for the fire's location. The best guess is compared against the ideal, previously known fireline, and found to be consistently accurate provided enough sensor nodes are distributed throughout the region. Further analysis still needs to be performed to determine the ideal number of sensor nodes required for any geographic area while maintaining accurate fire and outer fireline location. Even with the needed future analysis, the current results provide a strong base by which an argument can be made for the effectiveness and feasibility of such a system.
Enderle, Justin, "Wireless Sensor Network Wildfire Detection System" (2007). Opportunities for Undergraduate Research Experience Program (OURE). 188.